| Abstrakt: |
In robotic abrasive blasting operations, achieving the required surface roughness and cleanliness relies heavily on manually pre-set operational parameters. These parameters, such as stand-off distance, blasting angle, inlet air pressure, abrasive flow rate, and particle size, are set based on operator experience. However, given that optimal values for these parameters vary depending on specific blasting conditions and surface requirements, the pre-set values often fail to deliver both the desired surface quality and productivity. To address these challenges, we propose a solution employing a set of proxy models and a model predictive control system to achieve optimal operational outputs. For energy efficiency and productivity maximization, we utilize computational fluid dynamics with a multiphase flow solver to determine the optimal stand-off and offset distances, thus controlling the blasting nozzle effectively. Additionally, we build a data-driven model to obtain the optimal reference set point of air pressure at the sensor location based on the desired surface roughness. Moreover, we develop a dynamic process model to link the inlet air pressure and abrasive flow rate to the air pressure at the sensor location, which facilitates the development of a model-based control system. By applying this comprehensive approach, we evaluate and optimize the control and operational parameters to achieve the desired surface roughness and productivity targets. To validate the effectiveness of our newly developed control system and models, extensive tests are conducted both virtually (in silico) and on-site. The results confirm the stability and accuracy of the control system. In on-site reliability tests, the automated blasting system successfully delivers the desired surface roughness with a relative error of less than 5 % and a remarkable 30–50 % improvement in productivity. Overall, our innovative control system and models offer a reliable and efficient solution for robotic abrasive blasting, ensuring consistent surface quality and productivity enhancements. |
Full Text from ScienceDirect